V.G. Lyubov¹, I.G. Gattaullin²
¹Republican Clinical Oncology Dispensary named after Prof. M.Z. Sigal, Kazan
²KSMA — branch campus of the FSBEIFPE RMACPE MOH Russia, Kazan
Lyubov V.G. — Head of Day Hospital No. 9, Naberezhnye Chelny Branch of the Republican Clinical Oncology Dispensary named after Professor M.Z. Sigal
29 Sibirsky trakt St., 420029 Kazan, Russian Federation, e-mail: lyubov-vitalik@mail.ru, ORCID ID: 0009-0006-5356-4506
Abstract. Management of patients with persistently elevated prostate-specific antigen (PSA) and negative initial prostate biopsy remains a diagnostic challenge due to the risk of missed clinically significant cancer and unnecessary repeat biopsies.
The purpose — to review and systematize current evidence on the integration of multiparametric MRI, liquid biomarkers, and artificial intelligence (AI) to optimize the diagnostic algorithm in this patient population.
Material and methods. A literature search was performed in PubMed, MEDLINE, Scopus, Web of Science, Cochrane Library, CyberLeninka, and RSCI for the period 2020-2025. Keywords included: prostate cancer, negative biopsy, persistent PSA, mpMRI, PI-RADS, artificial intelligence, biomarkers. After screening, 50 sources were included.
Results. Three key strategies were identified: mpMRI with PI-RADS v2.1, novel biomarkers (PCA3, PHI, 4Kscore), and AI (deep learning, radiomics). Integration of these approaches can reduce unnecessary repeat biopsies by 30-60%. AI-based models allow combined analysis of clinical, imaging, and molecular data to provide individualized risk stratification.
Conclusions. The most promising direction is the development of AI-driven integrated predictive models. A practical diagnostic algorithm is proposed. Further prospective validation in real-world clinical settings is needed.
Key words: prostate cancer, negative biopsy, persistent PSA, artificial intelligence, multiparametric MRI, biomarkers, diagnostic algorithm